The
advancement
of
Large
Language
Models
(LLMs)
has
led
to
their
widespread
use
across
a
broad
spectrum
tasks,
including
decision-making.
Prior
studies
have
compared
the
decision-making
abilities
LLMs
with
those
humans
from
psychological
perspective.
However,
these
not
always
properly
accounted
for
sensitivity
LLMs'
behavior
hyperparameters
and
variations
in
prompt.
In
this
study,
we
examine
performance
on
Horizon
task
studied
by
Binz
Schulz
(2023),
analyzing
how
respond
prompts
hyperparameters.
By
experimenting
three
OpenAI
language
models
possessing
different
capabilities,
observe
that
fluctuate
based
input
temperature
settings.
Contrary
previous
findings,
display
human-like
exploration–exploitation
tradeoff
after
simple
adjustments
Journal of Computer Assisted Learning,
Год журнала:
2024,
Номер
40(5), С. 2369 - 2384
Опубликована: Июнь 30, 2024
Abstract
Background
With
the
development
of
artificial
intelligence
(AI)
technology,
generative
AI
has
been
widely
used
in
field
education
and
represents
a
groundbreaking
shift
overcoming
constraints
time
space
within
educational
activities.
However,
previous
literature
not
paid
enough
attention
to
AI‐involved
teaching
patterns,
it
is
necessary
evaluate
effects
this
learning
pattern.
Objective
s
Based
on
social
presence
theory
community
inquiry
model,
main
purpose
study
whether
how
interaction
frequency
with
chatbots
(IFC)
affects
people's
autonomy
(LA)
under
two
preferences:
knowledge
acquisition
virtual
companionship,
(SP)
plays
mediating
role.
Methods
The
1‐year
longitudinal
was
designed
be
conducted
from
May
2022
2023
included
three
rounds
surveys
1155
undergraduate
students
their
use
robots
for
learning.
Results
Conclusions
For
learners
preferring
no
direct
correlation
found
between
IFC
LA.
SP
acted
as
factor,
enhancing
LA
through
increased
chatbot
interactions.
This
suggests
that
while
interactions
may
directly
influence
LA,
resulting
can
foster
it.
Conversely,
favouring
acquisition,
higher
negatively
impacted
both
Despite
this,
strong
sense
consistently
correlated
positively
indicating
could
offset
some
negative
frequent
use.
Cyberpsychology Behavior and Social Networking,
Год журнала:
2025,
Номер
unknown
Опубликована: Фев. 5, 2025
Recent
research
has
investigated
the
connection
between
artificial
intelligence
(AI)
utilization
and
feelings
of
loneliness,
yielding
inconsistent
outcomes.
This
meta-analysis
aims
to
clarify
this
relationship
by
synthesizing
data
from
47
relevant
studies
across
21
publications.
Findings
indicate
a
generally
significant
positive
correlation
AI
use
loneliness
(r
=
0.163,
p
<
0.05).
Specifically,
interactions
with
physically
embodied
are
marginally
significantly
associated
decreased
-0.266,
0.088),
whereas
engagement
disembodied
is
linked
increased
0.352,
0.001).
Among
older
adults
(aged
60
above),
positively
0.001),
while
no
observed
0.039,
0.659)
in
younger
individuals
35
below).
Furthermore,
incorporating
attitudes
toward
AI,
study
reveals
that
influence
exacerbating
outweighs
reverse
impact,
although
both
directions
show
relationships.
These
results
enhance
understanding
how
usage
relates
provide
practical
insights
for
addressing
through
technologies.
Social Media + Society,
Год журнала:
2025,
Номер
11(1)
Опубликована: Янв. 1, 2025
Generative
chatbots
based
on
artificial
intelligence
technology
have
become
an
essential
channel
for
people
to
obtain
health
information.
They
provide
not
only
comprehensive
information
but
also
real-time
virtual
companionship.
However,
the
provided
by
AI
may
be
completely
accurate.
Employing
a
3
×
2
experimental
design,
research
examines
effects
of
interaction
types
with
AI-generated
content
(AIGC),
specifically
under
companionship
and
knowledge
acquisition
scenarios,
willingness
share
health-related
rumors.
In
addition,
it
explores
impact
nature
rumors
(fear
vs
hope)
role
altruistic
tendencies
in
this
context.
The
results
show
that
are
more
willing
situation.
Fear-type
can
stimulate
people’s
than
hope-type
Altruism
plays
moderating
role,
increasing
scenario
companionship,
while
decreasing
acquisition.
These
findings
support
Kelley’s
three-dimensional
attribution
theory
negativity
bias
theory,
extend
these
field
human–computer
interaction.
study
help
understand
rumor
spreading
mechanism
context
theoretical
improvement
chatbots.
Journal of Consumer Behaviour,
Год журнала:
2025,
Номер
unknown
Опубликована: Март 28, 2025
ABSTRACT
AI
assistants
are
transforming
the
business
landscape
by
revolutionizing
customer
service,
sales,
and
marketing.
This
study
investigates
how
consumer
ownership—defined
as
psychological
sense
of
ownership
over
products—affects
adoption
intentions
assistants.
Drawing
on
theory
commitment,
we
find
that
significantly
increases
intention
fostering
a
stronger
commitment
to
product.
In
series
three
studies,
demonstrate
real
(Study
1,
N
=
120,
Survey)
perceived
2,
200,
Experiment)
both
enhance
intentions.
Furthermore,
Study
3
(
reveals
impact
is
moderated
type
assistant.
Specifically,
effect
for
functional
but
disappears
companion
research
provides
valuable
strategies
businesses
increase
adoption.
British Journal of Management,
Год журнала:
2024,
Номер
36(1), С. 91 - 109
Опубликована: Май 14, 2024
Abstract
This
research
analyses
managers’
perceptions
of
the
multiple
types
artificial
intelligence
(AI)
required
at
each
stage
business‐to‐business
(B2B)
service
recovery
journey
for
successful
human–AI
collaboration
in
this
context.
Study
1
is
an
exploratory
study
that
identifies
main
stages
a
B2B
based
on
and
corresponding
roles
stage.
2
provides
empirical
examination
proposed
theoretical
framework
to
identify
specific
by
AI
enhance
performance
recovery,
perceptions.
Our
findings
show
prediction
benefits
from
collaborations
involving
processing‐speed
visual‐spatial
AI.
The
detection
requires
logic‐mathematical,
social
social,
verbal‐linguistic
post‐recovery
calls
